Research On IOs Application Security Evaluation Method Based On Mixed Features

With the rapid development of science and technology,the popularization of various intelligent terminal devices,and the change of people's lifestyle,iOS system as one of the most popular operating systems,security threats become more prominent.Due to the closed source nature of iOS system,its security research has certain limitations.This article intends to combine different detection methods to obtain different features,and uses different models to extract the security features of the application program for the feature evaluation.The main work and achievements of this paper are as follows:1)Study and analyze the security mechanism,system architecture,and operating mechanism of the iOS system;organize the typical malicious applications that appear in the current iOS system;analyze the current detection methods commonly used by application programs and compare the respective advantages and disadvantages.2)Putting forward the softmax classification model,which uses the static SS_API list to scan the static analysis feature vectors obtained by disassembling the iOS application code,and the network data analysis feature vectors to construct the SF feature as input for application.Program security threat classification.3)Analyze sensitive API calls to get their behavioral call sequences as feature inputs,optimize convolutional neural networks,and use flat long convolution kernels for convolutional calculations.4)An application security assessment model is proposed,and various types of static features,dynamic features,network data analysis features,and sensitive API call sequences are quantitatively analyzed to test and analyze the results.Make up for the inherent flaws of a single analysis.